Recombinant DNA techniques will be applied to three inter-related problems in evolutionary genetics: inferring population structure, detecting founder and bottleneck effects, and constructing phylogenies for closely related taxa. Population structure will be inferred by simultaneously studying maternally (mitochondrial DNA), paternally (Y-chromosomal DNA), and bisexually (X and autosomal DNA) inherited variants in the same population. Many aspects of population structure can be estimated with unisexually inherited markers that were previously difficult or even theoretically impossible to estimate. The amount and distribution of genetic variation observed in a species is a function of current population structure and past founder and bottleneck events in which the species underwent episodes of small population size. The detection of such founder and bottleneck events will be accomplished through contrasting restriction site polymorphism and evolutionary rates of mitochondrial DNA versus bisexually inherited ribosomal DNA and allozymes. Theory predicts a differential sensitivity to such historical events for haploid-maternal versus diploid-bisexual genetic systems. The historical information encoded in the DNA also extends back beyond the existence of the species itself, and hence phylogenies of closely related species will be constructed using this information. New statistical techniques for estimating and testing phylogenies will be developed. Closely related species are chosen because they are nearest to the speciation process. Various theories have proposed a role for population structure, founder events, and bottleneck effects in the process of speciation. Predictions from these theories will be tested by overlaying the phylogenetic reconstructions upon the information gathered concerning population structure and founder and/or bottleneck effects. This proposal is problem-oriented rather than organism-oriented. Organisms are chosen for which much ancillary information exists relating to at least one of the three specific aims, but for which important evolutionary questions remain to be answered. By choosing an appropriate array of organisms, we are simultaneously checking the validity of our inference procedure while expanding our knowledge of each group and testing basic predictions of speciation theory.